The Effect of Ram Pressure on the Molecular
The effect of ram pressure on the molecular gasof galaxies: three case studies in the Virgo cluster: three case studies in the Virgo cluster Bumhyun Lee, Aeree Chung, Stephanie Tonnesen, Jeffrey Kenney, O. Ivy Wong, B. Vollmer, Glen Petitpas, Hugh Crowl, Jacqueline van Gorkom
To cite this version:
Bumhyun Lee, Aeree Chung, Stephanie Tonnesen, Jeffrey Kenney, O. Ivy Wong, et al.. The effectof ram pressure on the molecular gas of galaxies: three case studies in the Virgo cluster: three case studies in the Virgo cluster. Monthly Notices of the Royal Astronomical Society, Oxford University Press (OUP): Policy P - Oxford Open Option A, 2017, 466 (2), pp.1382-1398. 10.1093/mnras/stw3162. hal-03161862
HAL Id: hal-03161862 https://hal.archives-ouvertes.fr/hal-03161862 Submitted on 17 Mar 2021
HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. MNRAS 466, 1382–1398 (2017) doi:10.1093/mnras/stw3162 Advance Access publication 2016 December 7
The effect of ram pressure on the molecular gas of galaxies: three case studies in the Virgo cluster
Bumhyun Lee,1‹ Aeree Chung,1,2,3‹ Stephanie Tonnesen,4 Jeffrey D. P. Kenney,5 O. Ivy Wong,6 B. Vollmer,7 Glen R. Petitpas,8 Hugh H. Crowl9 and Jacqueline van Gorkom10 1Department of Astronomy, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 03722, Korea 2Yonsei University Observatory, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 03722, Korea 3
Joint ALMA Observatory, Alonso de Cordova´ 3107 Vitacura, Santiago, Chile Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021 4Carnegie Observatories, 813 Santa Barbara St, Pasadena, CA 91101, USA 5Yale University Astronomy Department, PO Box 208101, New Haven, CT 06520-8101, USA 6International Centre for Radio Astronomy Research, The University of Western Australia M468, 35 Stirling Highway, Crawley, WA 6009, Australia 7CDS, Observatoire astronomique de Strasbourg, Universite´ de Strasbourg, CNRS, UMR 7550, 11 rue de l’Universite,´ F-67000 Strasbourg, France 8Harvard–Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138, USA 9Division of Science and Mathematics, Bennington College, 1 College Drive, Bennington, VT 05201, USA 10Department of Astronomy, Columbia University, Mail Code 5246, 550 W 120th St, New York, NY 10027, USA
Accepted 2016 December 2. Received 2016 December 2; in original form 2016 July 16
ABSTRACT We present 12CO (2–1) data of three Virgo spirals – NGC 4330, NGC 4402 and NGC 4522 obtained using the Submillimeter Array. These three galaxies show clear evidence of ram pressure stripping due to the cluster medium as found in previous H I imaging studies. Using the high-resolution CO data, we investigate how the properties of the inner molecular gas disc change while a galaxy is undergoing H I stripping in the cluster. At given sensitivity limits, we do not find any clear signs of molecular gas stripping. However, both its morphology and kinematics appear to be quite disturbed as those of H I. Morphological peculiarities present in the molecular and atomic gas are closely related with each other, suggesting that the molecular gas can be also affected by strong intracluster medium (ICM) pressure even if it is not stripped. CO is found to be modestly enhanced along the upstream sides in these galaxies, which may change the local star formation activity in the disc. Indeed, the distribution of Hα emission, a tracer of recent star formation, well coincides with that of the molecular gas, revealing enhancements near the local CO peak or along the CO compression. FUV and Hα share some properties in common, but FUV is always more extended than CO/Hα in the three galaxies, implying that the star-forming disc is rapidly shrinking as the molecular gas properties have changed. We discuss how ICM pressure affects dense molecular gas and hence star formation properties while diffuse atomic gas is being removed from a galaxy. Key words: ISM: molecules – galaxies: clusters: intracluster medium – galaxies: evolution – galaxies: ISM – galaxies: spiral.
deficient in H I compared to field galaxies based on their single- 1 INTRODUCTION dish observations. In addition, Giovanelli & Haynes (1985)have Since Gunn & Gott (1972) suggested that a galaxy might lose found that the H I content is well correlated with the location of its interstellar medium (ISM) by interacting with the intracluster galaxies in a sense that H I is more deficient at smaller distances medium (ICM) in the cluster environment, much evidence for ram from the cluster centre. Then the H I synthesis imaging studies such pressure stripping has been found to date. In early days, Davies & as Warmels (1988a,b) and Cayatte et al. (1990) have shown that H I- Lewis (1973) have shown that Virgo galaxies are generally more deficient galaxies near the cluster centre have small H I-to-optical extents, truncated within the stellar disc in many cases. In a more recent high-resolution H I imaging study of ∼50 selected Virgo E-mail: [email protected] (BL); [email protected] (AC) galaxies by Chung et al. (2009), a number of galaxies have been
C 2016 The Authors Published by Oxford University Press on behalf of the Royal Astronomical Society Molecular gas properties of Virgo spirals 1383 found with various scales of extraplanar H I gas or long H I tails, Therefore, in order to get a deeper understanding on how galax- indicating that ram pressure stripping is indeed acting in the cluster ies become passive after H I stripping, not only the possibility of environment. stripping but also the detailed properties of molecular gas inside a As a galaxy loses its ISM, the star formation rate is expected to galaxy under strong ICM pressure need to be probed. Hence, using be suppressed, which agrees well with the observations. As a good the Submillimeter Array (SMA),1 we have taken high-resolution example, Koopmann & Kenney (2004a) find that the massive star 12CO (2–1) imaging data of a subsample of Virgo galaxies that formation rate of Virgo spirals is lower than their field counterpart have lost atomic gas significantly by ICM pressure. Some prelim- by a factor of 2.5 on average. Koopmann & Kenney (2004b) also inary results are published in Lee & Chung (2015),andinthis show that many Virgo spirals have a smaller Hα extent compared work, we present more complete analysis on both morphology and to their stellar disc, reflecting that the process truncating H I discs kinematics of three Virgo galaxies based on the SMA data. may also be acting on the star-forming disc and Hα emission. This paper is organized as follows. We introduce the sample in On the other hand, molecular gas is unlikely to be as easily Section 2. Details of observations and data reduction procedure stripped as atomic hydrogen, since it is more tightly bound to the are provided in Section 3. In Section 4, we present the SMA data, galactic centre and the density is higher. In fact, most previous stud- describing the CO morphology and kinematics. In Section 5, we Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021 ies find that the molecular gas mass of the cluster population is not compare the CO and other wavelength data to discuss how the significantly different from that of field galaxies (Stark et al. 1986; molecular gas in these galaxies has been affected by the ICM pres- Kenney & Young 1989). In addition, more recent studies are finding sure. In Section 6, we summarize the results and conclude. clumpy dust features in the upstream side of H I gas stripped galax- A distance of 16 Mpc (1 arcsec ∼78 pc) to Virgo cluster is adopted ies, which are likely to be surviving dense clouds that are unveiled in this work (Yasuda, Fukugita & Okamura 1997). after diffuse atomic gas is removed (Crowl et al. 2005;Abramson & Kenney 2014; Kenney, Abramson & Bravo-Alfaro 2015). On the other hand, however, the opposite results have also been 2 SAMPLE GALAXIES reported. Rengarajan & Iyengar (1992) find that H2 mass normalized The sample for the SMA observations has been selected from by the dynamical mass of galaxies tends to get larger with increasing the VLA (Very Large Array)2 Imaging study of Virgo galaxies clustercentric distance, which supports that the molecular gas can be in Atomic gas (VIVA) by Chung et al. (2009). The VIVA is a high- also deficient in the cluster environment. More recently, Boselli et al. resolution H I imaging study of 53 late-type galaxies that are located (2014) show that H I-deficient galaxies in the cluster environment throughout the Virgo cluster from a high-density core region to low- tend to be also modestly deficient in molecular gas. density outskirts. Among the VIVA sample, we have selected three However, the molecular gas fraction to the optical luminosity of galaxies with clear evidence for active H I stripping, NGC 4330, spiral galaxies measured using 12CO (1–0) ranges quite widely for NGC 4402 and NGC 4522 (Fig. 1). Although these galaxies are all environments (Chung 2012). This implies that the molecular gas thought to have lost a similar fraction of H I gas, being deficient by fraction alone may not be a good tracer of molecular gas deficiency. a factor of 6 ∼ 7 compared to their field counterparts (Table 1), they Due to the scatter in this relation and opposing observational results, show distinct properties in their H I morphology. it is still arguable whether molecular gas can be affected by the ICM NGC 4330 is truncated in H I within one side of the stellar disc, I pressure in a similar way to H gas, and it is then still puzzling why while it reveals a long H I tail on the opposite side as if the H I disc star formation appears to be quenched in H I stripped galaxies if is pushed to the tail side (Chung et al. 2009;Abramsonetal.2011). molecular gas, which is the more direct ingredient for star formation, The location and the H I morphology suggest that this galaxy is is not deficient. a recent arrival, entering the high-density region for the first time No clear evidence for molecular gas stripping yet low star forma- (Chung et al. 2007), and this galaxy will reach the peak pressure tion activities in cluster galaxies may imply distinct molecular gas after 100 Myr, based on the simulation of Vollmer (2009). Mean- properties in high-density environments as supported by some pre- while, NGC 4402 has been experiencing strong ICM pressure in the vious observations. For example, Kenney et al. (1990) have found last ∼150−250 Myr (Abramson & Kenney 2014), currently cross- a highly asymmetric CO morphology in an H I deficient Virgo clus- ing the core region (Crowl et al. 2005). Lastly, NGC 4522 is farther ter galaxy. Vollmer et al. (2008) also have found very peculiar away from the cluster centre compared to NGC 4402 in projection, I CO distributions in some of Virgo spirals undergoing active H but its H I morphology is also suggestive of active ram pressure stripping. These studies indicate that the molecular gas can be po- stripping as NGC 4402. Vollmer et al. (2006) show in their simula- tentially disturbed by strong ICM pressure whether it is stripped or tions that it has been at least 50 Myr, since this galaxy experienced not. quite strong ICM pressure. This is likely due to turbulence in the Particularly along the side experiencing ICM pressure, it has ICM that could have been caused by merging of M49 group to the been suggested that interstellar gas including the molecular phase main cluster (Kenney, van Gorkom & Vollmer 2004). can be pushed up against the centre of a galaxy. This leads to ISM In spite of subtle differences in their H I morphologies, the mean compression and increasing H2 formation; hence, it locally triggers star formation quenching time-scale, i.e. how long ago a galaxy intensive star formation (Fujita & Nagashima 1999; Kronberger stopped forming stars, is not significantly different among our sam- et al. 2008; Merluzzi et al. 2013; Henderson & Bekki 2016). Indeed, ple by ranging from 100 to 300 Myr (Crowl & Kenney 2008; ISM compression with high molecular fraction in the upstream side is seen in a number of galaxies experiencing ICM pressure (Vollmer et al. 2012b; Nehlig, Vollmer & Braine 2016), which are 1 The SMA is a joint project between the Smithsonian Astrophysical Obser- often accompanied by an enhancement of star formation. Molecular vatory and the Academia Sinica Institute of Astronomy and Astrophysics gas enhancement among the Virgo cluster members has been also and is funded by the Smithsonian Institution and the Academia Sinica. recently reported by Mok et al. (2016). These observations clearly 2 The National Radio Astronomy Observatory is a facility of the National show that molecular gas properties and hence star formation activity Science Foundation operated under cooperative agreement by Associated within a stellar disc can be affected by ram pressure. Universities, Inc.
MNRAS 466, 1382–1398 (2017) 1384 B. Lee et al.
Abramson et al. 2011). Therefore, the detailed CO data of these three cases should enable us to probe how molecular gas proper- ties and thus star formation activities are modified by strong ICM pressure. In addition, by comparing our data with the CO data of NGC 4569, which is known to have crossed the cluster a while ago (Fig. 1, and has a star formation quenching time ∼300 Myr; Vollmer et al. 2004; Crowl & Kenney 2008), we will probe how molecular gas properties evolve with time, during the first infall to the cluster and under strong ICM pressure, then after core crossing. For NGC 4569, which is not included in our SMA sample, we make use of the CO data from the HERACLES survey (Leroy et al. 2009). The general properties of our SMA sample are summarized in Table 1. Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021 3 OBSERVATIONS AND DATA REDUCTION The SMA is a radio interferometer with eight antenna elements, 6 m each in diameter. It is located in Mauna Kea at the elevation of 4080 m above the sea level. Our SMA observations were done in March 2010 and March 2011 in the subcompact configuration. Among eight antennas in total, only seven antennas were available for our observations in both years. The total bandwidth of 4 GHz is composed of 48 chunks that are overlapped by ∼20 MHz. Each chunk was con- figured with 128 channels, each of which is 0.8125 MHz width or 1.1 km s−1 at the rest frequency of 12CO (2–1), 230.538 GHz. 12CO for each target was placed on the upper sideband so that 13CO (2–1) (νrest = 220.398 GHz) and C18O (2–1) (νrest = 219.560 GHz) fre- Figure 1. The locations of our sample are shown on the ROSAT X-ray map quencies were covered simultaneously. The lower sideband where (blue contours; Bohringer¨ et al. 1994). Red ellipses represent the position 13CO was included was separated with the upper sideband by angle and D25 × 5 of the sample in the B band. NGC 4569 that we have 10 GHz. not observed using the SMA but is included in our discussion as a good The primary beam of the SMA is ∼54 arcsec at 12CO (2–1) representative of a galaxy at post-peak pressure is shown in green. The rest frequency. Aiming to cover at least half the stellar disc, we sample galaxies, located at 0.4–1 Mpc from M87, adopting a Virgo distance mosaicked 3–5 points depending on the optical size of individual of 16 Mpc (Yasuda, Fukugita & Okamura 1997), make a nice sequence of galaxies as shown in Fig. 2. In the case of NGC 4522, one additional ram pressure stripping from early stage, close to peak pressure and post-peak α pressure in their orbit. field in the south-west was included to cover the extraplanar H and H I gas. The total integration time ranges from ∼2to∼5.5hper field depending on the weather conditions, yet we aimed to achieve a uniform sensitivity for each galaxy. Details of the SMA observations are summarized in Table 2.
Table 1. General information of sample galaxiesa.
Galaxy NGC 4330 NGC 4402 NGC 4522
Right ascension (J2000) 12h23m17s.012h26m07s.612h33m39s.7 Declination (J2000) +11◦2203. 5 +13◦0647. 4 +09◦1030. 2 Morphological type Sc Sb SBc Inclination (◦)798279 Position angle (◦)608935 −1 b Vrad(km s ) 1565 232 2328 D25(arcmin) 2.29 3.55 3.47 Total apparent B-band magnitude 12.02 12.05 11.86 9 c Total K-band luminosity (10 L,K) 6.58 21.30 5.64 M 8 d H I (10 M) 4.45 3.70 3.40 ef d,e d H I 0.80 0.74 0.86 ◦ d dM87( ) 2.1 1.4 3.3 aGeneral information of the sample galaxies from Paturel et al. (2003, HyperLeda, http://leda.univ-lyon1.fr/). bcf. the Virgo mean ∼1100 km s−1 (Mei et al. 2007). c × 10 × 11 Skrutskie et al. (2006), cf. Milky Way: 8.24 10 L, K (Drimmel & Spergel 2001), M31: 1.29 10 L, K (Barmby et al. 2006). d the VIVA study (Chung et al.2009). e def =log −log ,wherelog is the mean H I surface density of field galaxies (Haynes & Giovanelli 1984), H I H I,all H I,obs H I,all I and log H I,obs is the mean H surface density of an observed galaxy (Chung et al. 2009). In this work, morphology independent deficiency has been adopted as Chung et al. (2009).
MNRAS 466, 1382–1398 (2017) Molecular gas properties of Virgo spirals 1385
Figure 2. The H I distribution of NGC 4330, NGC 4402 and NGC 4522 (from left to right) is shown in blue contours overlaid on the Digitized Sky Survey Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021 2 (DSS2, https://archive.stsci.edu/dss/index.html) red image. The red cross indicates the stellar disc centre of each galaxy estimated from Spitzer 3.6 µm data (Salo et al. 2015), and the thin red circles represent the SMA observation points, each of which corresponds to the size of the primary beam at 230 GHz (≈54 arcsec).
Table 2. Observation parameters.
Galaxy NGC 4330 NGC 4402 NGC 4522
Observation date 2011 Mar 02, 03 2010 Mar 20, 21, 27 2011 Feb 28, Mar 01 Synthesized beam (arcsec) 6.35 × 4.47 7.21 × 3.89 6.62 × 3.92 Position angle (◦) −28.4 73.4 −26.4 Spectral resolution (km s−1) 5.0 5.0 5.0 Integration time per point (h) 2.1 5.5 2.6 rms per channel (mJy beam−1) 12CO J = 2 − 1 35.2 16.3 32.7 13CO J = 2 − 1a 29.7 13.0 29.9 Bandpass calibrators 0854+201, 1751+096 0854+201 1751+096, 3c279 Flux calibrators titan, vesta mars, titan mwc349a, vesta Gain calibrators 3c273, 3c279 3c273, 3c279 3c273, 3c279 Note. a13CO (2–1) emission is detected only in NGC 4402.
Flux, gain and bandpass calibrations were done using the MIR the receding and the approaching side, adopting the definition of software (Qi 2012). The calibrators used for our observations are Rhee & van Albada (1996): listed in Table 2. After the calibration, the data had been analysed W = V 20 per cent − V 20 per cent, using the MIRIAD. The continuum has been subtracted using UVLIN 20 high low (2) by applying a linear fit to the uv-data in line-free channels selected W = V 50 per cent − V 50 per cent. based on the H I emission of each galaxy. The imaging and mosaick- 50 high low (3) ing were done using INVERT. In order to maximize the sensitivity The CO velocity is determined following Verheijen (1997): while keeping optimal resolution, the robust was set to 0.5 (for NGC V = . V 20 per cent + V 50 per cent 4402) or 1 (for NGC 4330 and NGC 4522). The channel width in sys 0 25 low low − the final cube was regridded to 5 km s 1, which is comparable to 50 20 H I data (VIVA; Chung et al. 2009). We have detected 12CO (2–1) + V per cent + V per cent . high high (4) in all three galaxies, while 13CO (2–1) has been detected only in NGC 4402, which will be presented in a separate paper (Lee et al. Following the recipe from Solomon et al. (1997), 12CO (2–1) in preparation). line luminosity is measured as follows: L − = . × 7S ν−2D2 + −3, CO(2 1) 3 25 10 CO obs L(1 z) (5) 4 RESULTS −1 2 in K kms pc ,whereSCO is an integrated total 12CO (2–1) flux −1 ν In this section, we present the results of our SMA observations. The in Jy km s , DL is the luminosity distance in Mpc, obs is the 12CO (2–1) flux and its uncertainty are measured in Jy km s−1 as observation frequency in GHz and z is the redshift. Adopting a ≈ follows: typical 12CO (2–1)/(1–0) ratio of nearby galaxies (R21 0.8; Leroy 1/2 et al. 2009) and using the conversion relation of normal spirals 2 −2 −1 −1 SCO = FCO × V ± σ × V , (1) (αCO = 3.2 M pc (K km s ) ; Strong & Mattox 1996), the molecular gas mass can be calculated by where FCO is the total flux of CO emission in each channel, V αCO is the channel separation of the final cube (5 km s−1)andσ is the M = L (2−1), (6) H2 R CO rms of each channel outside CO emission. Continuous features 21 σ σ L above 2 –3 are considered as real signal in individual channels. where CO(2–1) is the luminosity of 12CO (2–1) transition. The CO linewidths are calculated using the velocities, where the flux global CO properties of the sample are summarized in Table 3. density corresponds to 20 per cent and 50 per cent of the peaks in On the left of Figs 3–5, the CO intensity map, velocity field and
MNRAS 466, 1382–1398 (2017) 1386 B. Lee et al.
Table 3. SMA CO properties of sample galaxies. (FUV) emission (Fig. 4e), which will be discussed more in detail in Section 5. The northern part above the major axis is measured to be NGC 4330 NGC 4402 NGC 4522 91 per cent of that of the south in flux. Although the CO peak agrees −1 well with the stellar disc centre, the inner CO disc also appears to W20 (km s ) 208 255 176 −1 be slightly more stretched towards the west, reflecting the outer CO W50 (km s ) 184 226 159 −1 Vsys (km s ) 1562 246 2326 disc. −1 SCO (Jy km s ) 182.22 ± 8.02 1400.76 ± 11.91 139.43 ± 4.77 The CO kinematics reveals the evidence for non-circular mo- M 8 a ± ± ± H2 (10 M) 1.19 0.05 8.83 0.08 0.88 0.03 tions as shown in Fig. 4(b). In the inner region, the isovelocity a −2 −1 −1 curves are highly skewed and not perfectly parallel with the minor Note. The CO-to-H2 conversion factor of 3.2 M pc (K km s ) is adopted from Strong & Mattox (1996). axis. The skewness in velocity must be the result of two distinct disc components in the inner ∼15 arcsec (∼1 kpc) being projected on the sky as seen in its PVD (Fig. 4c). The velocity gradient position–velocity diagram (PVD) with global and radial profiles in the outer part is quite different in the two sides of the disc along
are presented in (a), (b) and (c) for each galaxy. The overlays with Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021 the major axis. In the north-west quarter (in the downstream side), other wavelength data shown on the right-hand side of Figs 3–5 are the kinematics of modest CO bump is slightly deviated from the discussed in Section 5. main disc. It also has a distinct CO component with steep velocity gradient ∼ 4.1 NGC 4330 in the inner 15 arcsec. This likely indicates a nuclear bar or ring, although we favour a nuclear bar as this would also explain the As seen in Figs 3(a)–(c), the CO morphology of NGC 4330 is skewed isovelocity contours. In NGC 4402, Sofue et al. (2003a,b) highly asymmetric, with the south-west extent in the downstream also find a nuclear molecular disc traced by 12CO (1–0) in the side being about 75 per cent (40 arcsec versus 53 arcsec) of the inner ∼10 arcsec. We do not find any direct relevance of these north-east but only 68 per cent when the outermost 4 arcsec of the nuclear structures with ram pressure stripping. south-west disc, which is bent down to the south (indicated by arrow in Fig. 3a) is excluded (36 arcsec versus 53 arcsec). The central part, within 30 arcsec in diameter (∼2 kpc at the distance of Virgo), is also asymmetric due to the strong local peak in the north-east from 4.3 NGC 4522 the stellar disc centre. The distinct extents and surface densities are While the overall size of 12CO (2–1) disc in the SMA image is also clearly seen in radial profiles. In the end of the south-west disc consistent with the extent measured by a single dish for NGC 4330 of the downstream side, CO is found to be slightly bent (Fig. 3a). (Vollmer et al. 2012a) and NGC 4402 (Lee et al. in preparation), This bending is also found in many other wavelengths such as UV, our SMA data of NGC 4522 reveal only the inner ∼57 per cent α H and H I, but all are different in scale and angle from one another of its single-dish map in size (Vollmer et al. 2008). While Vollmer as further discussed in Section 5. In the case of CO, this bending et al. (2008) have detected CO along the extraplanar Hα and H I part is clumpy, almost identified as an independent blob or clump in both ends of the disc, we did not detect any such features in the (see the arrow in Fig. 3c). This CO clump corresponds to 2 per cent south-west in our SMA data due to the lack of sensitivity and/or ∼ × 6 of the total in flux ( 1.9 10 M, comparable to that of a large potentially due to the diffuse nature of gas in the outer molecular molecular cloud). The centre of the clump is off from the mid-plane disc. The north-east end was not covered in the SMA observations. ∼ ∼ of the main disc towards the south by 4arcsec( 312 pc). Therefore, we limit our discussion to only the inner CO disc for The CO kinematics also shows peculiar structures, especially this particular case in this section and we cite the IRAM single-dish along the south-west concentration and the end of the tail as shown data when it is needed for comparisons with other wavelength data in Fig. 3(b). The velocity gradient of the south-west is quite steep, in Section 5. while it is more slowly rising on the other side within the small radii As shown in Fig. 5, the CO extent is comparable in both sides −1 from the centre, reaching 72 km s on the approaching side but (27.4 arcsec in the north-east versus 27.6 arcsec in the south-west). −1 only 43 km s on the receding at 15 arcsec radii. In the south-west The inner CO disc of NGC 4522 is found to be curved in the opposite clump, the velocity gradient is inverted. The kinematical complexity waytotheouterCOandHI disc, i.e. to the south-east as the inner Hα is also clearly seen in the PVD (Fig. 3c). A distinct component of a disc. We find two features sticking out from the inner part, one from ∼ ∼ steep velocity gradient in the central 13 arcsec ( 1 kpc) is quite the end of north-east disc and another almost to the same direction noticeable. This may indicate a molecular ring or bar. In addition, but connected from the centre of the main disc, more to the south. ∼ the CO peak is off from the stellar disc centre by 270 pc (Fig. 3a), The north-east blob coincides with the morphology of the inner CO while the CO kinematic centre is more or less consistent with the disc in the IRAM image (Fig. 6c; Vollmer et al. 2008), which is stellar disc centre (Fig. 3c). smoothly connected to the disc farther extended in the north-east, coinciding with a dust loop in this region. While the downward bending coincides with the inner spiral structure including the dust 4.2 NGC 4402 feature (see Fig. A1 in Appendix A). AsshowninFig.4(a), the CO disc of NGC 4402 is slightly more The CO velocity structure within the main disc generally shows extended in the west (62 arcsec in the west versus 58 arcsec in that of a regularly rotating disc as seen in Figs 5(b) and (c). The the east), but the difference is subtle and not as significant as in velocity keeps rising up to ∼10 arcsec on both sides, then almost H I. What makes this case look highly asymmetric is the north- flattens out, and towards the end of the disc, the velocity rises west quarter of the downstream side (the west side of the disc). again. The velocity gradient along the north-east branch is overall In this region, a modest CO bump (indicated by arrow in Fig. 4a) consistent with that of the main disc on the same side. However, the is found as in H I (Fig. 4d). Meanwhile, the southern part of the gradient is smaller and the velocity rises more slowly compared to CO disc looks quite compressed along the enhanced far-ultraviolet the other side.
MNRAS 466, 1382–1398 (2017) Molecular gas properties of Virgo spirals 1387 Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021
Figure 3. NGC 4330: (a) 12CO (2–1) integrated intensity map (the 0th moment) in grey-scale with contours. Contour levels are 0.3, 1.5, 4, 8, 12, 16, 20 Jy beam−1 km s−1. The synthesized beam size is 6.35 arcsec × 4.47 arcsec (blue ellipse at the bottom right). The white cross indicates the stellar disc centre. (b) 12CO (2–1) velocity field map (the 1st moment). Velocity contours are drawn in every 10 km s−1 from 1470 to 1670 km s−1. The white cross again indicates the stellar disc centre. (c) Upper left: a position–velocity cut through the major axis integrated along the minor axis. Contour levels are 0.7, 1.4, 2.1, 2.8, 3.5 Jy beam−1. The CO clump is indicated by black arrow. Right: the global profile of 12CO (2–1). The CO velocity (1562 km s−1) is indicated with an arrow. Bottom: the gas surface densities (H I and H2) along the approaching side and the receding side on the right and the left, respectively. (d) An overlay of 12CO (2–1) (blue contours) on H I (red contours) and DSS2 red (black contours). Synthesized beam of the VLA and the SMA are shown in red and blue at the bottom left. The entire CO disc is located inside the optical disc. Unlike the stellar disc, however, the CO disc is found to be highly asymmetric as the atomic gas disc. (e) 12CO (2–1) (blue contours) is overlaid on FUV emission (black contours and grey-scale). FUV tail is also extended and bent towards the south-west as CO and H I gas tail. (f) 12CO (2–1) (blue contours) overlaid on the Hα emission (grey-scale). The overall bending shape coincides well between Hα and CO, while CO is not extended as much as Hα.
MNRAS 466, 1382–1398 (2017) 1388 B. Lee et al. Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021
Figure 4. NGC 4402: (a) 12CO (2–1) integrated intensity map (the 0th moment) in grey-scale with contours. Contour levels are 1, 5, 10, 30, 50, 70, 90 Jy beam−1 km s−1. The synthesized beam size is 7.21 arcsec × 3.89 arcsec (blue ellipse at the bottom right). The white cross indicates the stellar disc centre. A modest CO bump is indicated by black arrow. (b) 12CO (2–1) velocity field map (the 1st moment). Velocity contours are drawn in every 20 km s−1 from 120 to 360 km s−1. The white cross again indicates the stellar disc centre. (c) Upper left: a position–velocity cut through the major axis integrated along the minor axis. Contour levels are 0.5, 1.5, 3, 5, 7, 9, 11 Jy beam−1. Right: the global profile of 12CO (2–1). The CO velocity (246 km s−1) is indicated with an arrow. Bottom: the gas surface densities (H I and H2) along the approaching side and the receding side on the right and the left, respectively. (d) An overlay of 12CO (2–1) (blue contours) on H I (red contours) and DSS2 red (black contours). 12CO gas is well confined within the optical disc, while H I is pushed off outside from the stellar disc. The optical disc does not look disturbed, while both molecular and atomic gas components reveal asymmetry in a similar sense. (e) 12CO (2–1) (blue contours) is overlaid on FUV emission (black contours and grey-scale). FUV is found to be enhanced along the CO compression. (f) 12CO (2–1) (blue contours) overlaid on the Hα emission (grey-scale). The overall extent and morphology well coincide with each other.
MNRAS 466, 1382–1398 (2017) Molecular gas properties of Virgo spirals 1389 Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021
Figure 5. NGC 4522: (a) 12CO (2–1) integrated intensity map (the 0th moment) in grey-scale with contours. Contour levels are 0.3, 1.5, 4, 8, 12, 16, 20 Jy beam−1 km s−1. The synthesized beam size is 6.62 arcsec × 3.92 arcsec (blue ellipse at the bottom right). The white cross indicates the stellar disc centre. (b) 12CO (2–1) velocity field map (the 1st moment). Velocity contours are drawn in every 10 km s−1 from 2250 to 2400 km s−1. The white cross again indicates the stellar disc centre. (c) Upper left: a position–velocity cut through the major axis integrated along the minor axis. Contour levels are 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0 Jy beam−1. Right: the global profile of 12CO (2–1). The CO velocity (2326 km s−1) is indicated with an arrow. Bottom: the gas surface densities (H I and H2) along the approaching side and the receding side on the left and the right, respectively. (d) An overlay of 12CO (2–1) (blue contours) on H I (red contours) and DSS2 red (black contours). Our SMA data reveal only the inner part of the molecular gas disc, missing out the outer part that is detected by the IRAM (Vollmer et al. 2008) due to the lack in pointings and sensitivity. In Sections 4 and 5, we focus more on the morphology and kinematics revealed by the SMA. (e) 12CO (2–1) (blue contours) is overlaid on FUV emission (black contours and grey-scale). Enhanced FUV emission is found along the side on which the ICM wind is presumably acting. (f) 12CO (2–1) (blue contours) overlaid on the Hα emission (grey-scale). The inner CO disc is slightly bent towards south-east as the inner Hα disc.
MNRAS 466, 1382–1398 (2017) 1390 B. Lee et al. Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021
Figure 6. A composite map of FUV (blue; Gil de Paz et al. 2007), Hα (NGC 4330 and NGC 4569 from Gavazzi et al. 2003, NGC 4402 from Crowl et al. 2005, NGC 4522 from Koopmann, Kenney & Young 2001), optical (black; DSS2 red), H I (green; Chung et al. 2009) and 12CO (2–1) (white contours from our SMA data) of NGC 4330 (redshifted), NGC 4402 (blueshifted), NGC 4522 (redshifted), and NGC 4569 (blueshifted). The IRAM 12CO (2–1) data are shown in yellow contours. Contour levels are 3.7, 7.5 Jy beam−1 km s−1 for NGC 4330 (Vollmer et al. 2012a); 5, 20 Jy beam−1 km s−1 for NGC 4402 (Lee et al. in preparation); 1.7, 3.5, 5.3 Jy beam−1 km s−1 for NGC 4522 (Vollmer et al. 2008) and 14, 71, 106, 214, 498, 712 Jy beam−1 km s−1 for NGC 4569 (Leroy et al. 2009). The ICM wind direction deduced from H I morphology is shown with black arrows (Vollmer et al. 2004; Crowl et al. 2005; Abramson et al. 2011; Abramson & Kenney 2014). NGC 4330, NGC 4402 and NGC 4522 are at relatively early to active H I stripping stage due to the ICM, while NGC 4569 is thought to be a post-peak pressure case where some H I gas is falling back on the disc after core crossing.
5 DISCUSSION and 100 Myr, respectively (Kennicutt 1998). We also compare the radial gas surface density between atomic and molecular hydrogen. 5.1 Comparison with other wavelength data The H I column density is calculated using In this section, we compare 12CO (2–1) properties with other 18 N(H I) = 1.82 × 10 × TB V (7) wavelength data including H I (Chung et al. 2009), optical (DSS2 α −2 red), FUV (Gil de Paz et al. 2007)andH (NGC 4330 and NGC in cm ,where TBV is the integrated intensity (brightness tem- −1 4569 from Gavazzi et al. 2003,Abramsonetal.2011; NGC 4402 perature) in K kms (Walter et al. 2008). The H2 surface density is from Crowl et al. 2005; NGC 4522 from Koopmann, Kenney & estimated by the following relation: Young 2001). CO traces star-forming gas, and Hα and FUV emis- αCO sion are good indicators of star formation with time-scales of ∼20 N(H2) = I(CO) (8) R21
MNRAS 466, 1382–1398 (2017) Molecular gas properties of Virgo spirals 1391 in M pc−2,whereI(CO) is 12CO (2–1) integrated intensity, and As clearly seen, the molecular disc, which extends to only half α CO is the same as in equation (6). Considering the high inclina- the stellar disc or less, shows many properties in common with the R21 tion of all three galaxies in our sample, we utilize the strip integral morphologies in the other wavelengths, reflecting the impact of the method instead of ellipse fitting to derive radial surface density be- ICM pressure. Although there are subtle differences, this suggests cause ellipse fitting is usually inappropriate for edge-on galaxies that the molecular gas in this galaxy has been also affected in similar when the spatial resolution is limited (Warmels 1988a;Rhee&van ways by the same mechanism that is responsible for the peculiarities Albada 1996;Swaters1999). This method calculates strip integrals found in the other wavelengths. This indicates that ICM pressure by integrating H I column density perpendicularly to the major axis, can change molecular gas properties well inside the stellar disc. using Lucy’s (1974) iterative deconvolution method. Then the strip integrals of approaching and receding sides are deprojected sepa- rately to infer the face-on H I surface density (Kregel, van der Kruit 5.1.2 NGC 4402 & de Blok 2004). The surface density of both atomic and molec- The CO emission shows a number of properties in common with ular hydrogen is measured in M pc−2. The comparison between I the H gas (Fig. 4d). It is somewhat extended in the west, i.e. the Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021 atomic and molecular gas radial distributions is found in the bottom same side where the H I is more extended. Both phases show a on the left-hand column of Figs 3–5. Overlays between CO and bump in the west (indicated by arrow in Fig. 4a), although the scale other multiwavelength data are presented on the right-hand column is distinct because the H I bump is visible in extraplanar gas. H I is of Figs 3–5 and Fig. 6. truncated within the mid-plane of the stellar disc on both sides, with For the reference, the IRAM 12CO (2–1) maps are also overlaid a sharp cut-off in the east to south along the upstream. Meanwhile, in the composite map. The sensitivity of the IRAM cubes is ∼10 −1 the northern side is more extended with a short tail pointing the mK per 5 km s channel (the same channel width as ours), which north-west in the downstream side. The extent of the molecular gas is comparable to the SMA cubes or slightly better in the cases of is not significantly different in the receding and the approaching NGC 4330 and NGC 4522 when the different spatial resolutions side, yet its distribution is quite distinct. These differences between ∼ ∼ are taken into account ( 11 arcsec versus 5 arcsec for IRAM the two sides of the CO disc can be also clearly seen in the radial and SMA, respectively). For NGC 4402, the sensitivity of SMA velocity distributions (Fig. 4c). Note that the H I extent appears data is somewhat better than the IRAM data not only because of the to be almost twice as long as the CO in the west due to the tail sufficient integration time with the SMA but also because of the poor (Figs 4c and d). weather condition when the IRAM data were taken. For NGC 4330 The molecular gas disc is thicker in the east (especially third con- and NGC 4402, the IRAM and SMA maps are generally in good tour level from the outermost contour), reflecting the change in FUV agreement in morphology, considering the beam sizes, while we disc thickness across the disc. Although the overall morphology of are missing most of the extraplanar molecular of NGC 4522 in the FUV is quite different from that of CO, both the FUV enhancement SMA data due to the lack in coverage, sensitivity and/or potentially and CO compression are present in the east and south-east (Fig. 4e). due to the diffuse nature of molecular gas in the extragalactic space Hα is tightly correlated with CO in general, with a similar distribu- as further discussed in Section 5.1.3. tion and extent. Also, strong Hα knots are found near or within the local CO peaks (Fig. 4f). As with NGC 4330, the CO morphology is surprisingly similar to that of H I, strongly suggesting that the 5.1.1 NGC 4330 impact of the ICM has reached the inner interstellar gas disc. Asymmetry is ubiquitous in a range of wavelengths data except in the red optical image (DSS2 red). In Fig. 6(a), we see that the 5.1.3 NGC 4522 H I is truncated within the stellar disc in the north-east, where Hα and FUV emission reveal an upturn feature (Chung et al. 2009; With given sensitivity limit, we are missing the outer part of the Abramson et al. 2011). A hint of upturn is also found in the north- molecular gas disc in the SMA data for this case. As shown by east end of the upstream of radio continuum at 6 and 20 cm (Chung Vollmer et al. (2008), however, the overall morphology of CO is et al. 2009; Vollmer et al. 2012a). The single-dish 12CO (2–1) data quite similar to that of H I, revealing a significant fraction of molec- (Vollmer et al. 2012a) appear to be similar to the radio morphology, ular gas outside the stellar disc (Fig. 6c). This galaxy must be one also with some hint of upturn, although the extent of the molecular of the best examples found to date where extraplanar molecular gas gas is much less than that of H I. In our SMA data, we are missing is present. In addition to the sensitivity, the molecular gas could be some outermost features including the tip of the nose in the north- more diffuse in nature compared to the gas within the disc, which east. Instead, the SMA data clearly reveal very detailed structures of could make the detection more difficult. In fact, we recently have the inner CO disc such as the difference in scaleheight between the detected some extragalactic 13CO (1–0) in NGC 4522 using the north and the south along the major axis (up to ∼0.36 kpc, Fig. 3a), Atacama Large Millimeter Array (ALMA) – cycle 3. The nature which is not as clear as this in the previous single-dish data. of this extraplanar CO gas will be further analysed using 12/13CO The south-west side, i.e. the downstream of the galaxy where an ratio, and will be presented in a separate work. H I gas tail is present, is bent towards the south. On this side, we The inner CO disc is slightly bent down to the south as Hα also find FUV, Hα and radio continuum tail but the extents and emission. The strongest Hα blobs near the galactic centre are located bending angles are all different in various wavelengths (Fig. 6a). within the inner CO disc (Fig. 5f), but there is an offset between the The distribution becomes patchy in the outer CO disc as seen in inner FUV enhancement and the inner CO disc (Fig. 5e). However, both IRAM single-dish data (Vollmer et al. 2012a) as well as our this offset in the inner region might be simply due to the dust SMA data. Intriguingly, the CO blob towards the end is located in extinction effect. On the other hand, the offset between CO (the almost the same galactic radius as one of the distinct Hα blobs and IRAM data)/Hα and FUV at large radii is real since there is almost off from the mid-plane in the same direction as the tails of the other no dust beyond the CO extent. The extent of FUV is larger than the wavelengths (Fig. 6a). CO and Hα extents within the stellar disc by a factor of 1.5 (Fig. 6c).
MNRAS 466, 1382–1398 (2017) 1392 B. Lee et al. Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021
Figure 7. The PVD of each galaxy (black contours) is mirrored (red contours) and overlaid on the original one. These images are folded along the centre of the stellar disc. The kinematics of molecular gas in NGC 4330 and NGC 4402 is quite distinct in the receding and the approaching side. In the bottom row, the comparisons with the H I PVD (blue contours) are presented.
However, the FUV enhancement is roughly consistent with the CO (DSS2 red) of our SMA sample are presented. A rough direction of compression in the downstream side as in NGC 4402. This implies the ICM wind based on the H I morphology for each case is shown that there is significant change in the recent star formation activity with thick arrows. and the distribution of molecular gas in a very short time as further Secondly, the detailed CO structure of the sample shows good discussed in Section 5.3. connections with the H I features that have been pushed and/or stripped by the ICM pressure. In NGC 4330, the CO clump in the south-west end seems to have been pushed down to the south from 5.2 Impact of ram pressure on molecular gas the mid-plane of the stellar disc, similarly to its H I tail (Fig. 3d). In addition, similar kinematical structures are found in this side, i.e. Most previous studies using single-dish radio telescopes did not find in the downstream in both H I and CO (Fig. 7). For example, the a significant difference in CO luminosities between field galaxies CO gas reveals an inverted velocity gradient in the end of south- and cluster members (Stark et al. 1986; Kenney & Young 1989). west like H I, which shows an anomalous velocity component in the However, some individual galaxies in the cluster environment have same side (Abramson et al. 2011). This implies that dense molecular been reported to show very distinct CO distributions from those phase gas can be pushed by ram pressure in addition to the diffuse with normal H I content in the field (e.g. Kenney et al. 1990; Sofue atomic phase gas. Although the morphology is suggestive of gas et al. 2003c). More recently, Fumagalli et al. (2009) have shown that being pushed away, one problem with this possibility is that both the molecular gas surface density of galaxies that are H I deficient the H I tail and the CO clump in the downstream are found with tends to be low. These suggest that the molecular gas of cluster higher velocities than the main disc, i.e. the opposite direction of galaxies can be affected in some ways by ram pressure even if it is the system velocity of cluster centre. This is inconsistent with what not stripped. Indeed, our high-resolution CO data clearly show that is generally expected in the case of the gas pushed by the ICM as the molecular gas has been affected in a similar manner as diffuse found in other Virgo spirals (e.g. NGC 4522; Kenney, van Gorkom gas by ram pressure, and the impact of ICM winds reaches as &Vollmer2004). However, this is still feasible if the gas has been deeply as a few kpc from the galactic centre, as we further elaborate losing angular momentum in extraplanar space as suggested by below. Abramson et al. (2011). First, the CO morphology in our sample is all found to be highly Alternatively, the CO clump might be an example of a dense cloud asymmetric and disturbed. The inner molecular gas disc of many unveiled after the stripping of surrounding diffuse gas, which hence field galaxies does show some peculiarities to a certain degree, in- appears to stand alone, decoupled from the rest of the disc, similar cluding asymmetry (e.g. BIMA SONG survey; Helfer et al. 2003). to the ones seen as narrow dusty plumes found in Virgo galaxies However, the overall asymmetry found in the CO morphology such as NGC 4402 and NGC 4522 (Crowl et al. 2005;Abramson& of all three galaxies is similar to that of the H I morphology. In Kenney 2014; Kenney, Abramson & Bravo-Alfaro 2015)orNGC Figs 6(a)–(c), the composite maps of cool ISM contents (CO and 4921, a Coma spiral (Kenney et al. 2015). If this is an example of H I), star formation indicators (FUV and Hα) and the old stellar disc those decoupled clouds, it is the first time it has been seen clearly
MNRAS 466, 1382–1398 (2017) Molecular gas properties of Virgo spirals 1393 in CO. However, one aspect that makes this case particularly inter- significantly during and after severe H I stripping, while its detailed esting is its gas kinematics. Rather than being smoothly connected molecular gas properties may change. These changes to the CO gas from the inner disc, the velocity gradient at the location of the CO properties are consistent with those from Wong et al. (2014), who clump turns back. This velocity structure strongly suggests that this found enhanced levels of warm molecular gas that resulted from clump is the molecular gas that has been pushed off from the main additional ISM-heating due to ram pressure driven shocks in NGC disc. 4330, NGC 4402 and NGC 4522. In NGC 4402, both H I and CO are compressed from the south- east to south along the upstream. Meanwhile in the north-west, a modest CO bump is found to be pointing to the same direction as 5.3 Evolution of molecular gas and star formation activities the H I bump (Fig. 4d). The compressed CO in the upstream side by ram pressure reveals a steeper velocity gradient than the other side. Meanwhile, 5.3.1 Galactic disc the CO bump in the north-west downstream side (Fig. 4b) is found to have larger velocities than the main disc (Fig. 4c), i.e. closer to All three galaxies have lost H I gas significantly, containing the cluster centre. This is consistent with our expectations if the gas only <∼20 per cent of H I mass compared to field galaxies of Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021 is accelerated by ram pressure due to the ICM. similar size. The star formation quenching time-scale, τ q, mea- In NGC 4522, the overall CO morphology from the IRAM data sured based on the youngest stellar age at the location, where H I (Vollmer et al. 2008) is very similar to the H I morphology, with the gas is truncated (i.e. one way to estimate how long ago star for- extraplanar gas in the downstream side and compression in the up- mation was suppressed in a galaxy), ranges from a few tens Myr stream side (Fig. 6c). This strongly suggests that the molecular gas to a few hundreds Myr for our sample (Crowl & Kenney 2008; has been influenced by the same mechanism that affects the atomic Abramson et al. 2011). It is however comparable among the sam- gas, i.e. ram pressure due to the ICM wind. In the upstream side, ple within the uncertainties, and they share several properties in the interstellar gas can also get compressed by the ICM pressure, common in star formation activity in spite of subtle differences due which potentially results in the formation of molecules as inferred to the ICM wind angle and the projection on the sky as shown in by the excess of CO in this side (see also the example of NGC 4654; Fig. 6. Chung & Kim 2014). The spatial distributions of CO and Hα (a tracer of recent star AsshowninFig.7, the CO kinematics also suggests that the formation with a time-scale of <∼20 Myr; Kennicutt 1998)are ICM pressure can disturb the molecular gas, deep inside a galactic remarkably well correlated in the inner few kpc radius (Fig. 6). The potential well. In NGC 4330 and NGC 4402, the CO velocity gra- locations of local CO peaks are generally coincident with the Hα dients are quite distinct in the receding and the approaching side, an knots as shown by high-resolution data. The extents and overall asymmetry also observed in H I. In the outer part of NGC 4330 and morphology are also similar to each other as seen in the comparison NGC 4402 (the CO clump and the modest CO bump, respectively) with single-dish data, although the Hα blobs outside the old stellar where the molecular gas is off from the mid-plane, we also find the disc do not always have a CO counterpart (e.g. in both ends of NGC velocity gradient deviates from the overall gas flow of the gas disc. 4330’s FUV disc; Fig. 6a). The comparisons with the H I PVD is also intriguing (the bottom Meanwhile, FUV (a star formation tracer of a time-scale row of Fig. 7). In the upstream of NGC 4330 and NGC 4402, the of ∼100 Myr; Kennicutt 1998) is somewhat distinct from CO and velocity gradients of the molecular gas and atomic gas are similar Hα in extent and morphology, at least on one side, with FUV to each other, but they deviate in the downstream. In NGC 4522, being more extended than CO and Hα. Although the FUV disc the deviation starts at ∼0.8 kpc from the centre on both sides. This of many field galaxies is larger than their Hα disc (e.g. Leroy may indicate a delay in stripping of gas in distinct phases, i.e. more et al. 2008), these three galaxies are distinct from the field popula- diffuse gas dragged prior to relatively dense gas. tion in a sense that both FUV and Hα are highly truncated within Our results clearly show that the molecular gas in galaxies ex- the stellar disc. These observations indicate that star-forming discs periencing active H I stripping can become highly asymmetric, as have been shrunk considerably (by a factor of up to 1.5) while un- demonstrated by simulations (e.g. Hidaka & Sofue 2002; Vollmer dergoing strong ram pressure. Then the question is what causes the et al. 2008, 2012a). In addition, our high-resolution data indicate quenching of star formation, whether it is molecular gas stripping, that the influence of strong ICM pressure reaches quite deep inside the change of molecular gas properties, or both. of the galactic disc. Although the molecular gas may be pushed off As seen by the morphological/kinematical peculiarities of molec- from the stellar disc together with the atomic gas as in NGC 4522 ular gas in our sample, the ICM pressure is strong enough to strip (Vollmer et al. 2008), active ram pressure does not seem to always the H I gas in the outer disc or push the H I disc to one side, and cause molecular gas stripping (NGC 4330 and NGC 4402). Indeed, seems to be able to also affect the molecular gas at a few kpc radii NGC 4569, which is thought to have lost more than 90 per cent of of the stellar disc. This can cause a number of consequences. First, its H I gas while crossing the cluster core a few hundreds Myr ago in the outer disc where stars must have been formed mostly from (H I deficiency: 1.47, a projected distance of 1.7 deg; ∼0.5 Mpc to low-density interstellar gas, star formation activity will gradually M87, the bottom right of Fig. 6), has a comparable molecular gas decrease as the atomic gas is stripped. Even when quenching starts, fraction to optical luminosity of our SMA sample (NGC 4330, NGC some young stars from the last star formation event may be still ob- 4402 and NGC 4522). Adopting the H2 deficiency of Fumagalli et al. servable, and indeed, the FUV emission outside CO and Hα discs (2009), single-dish measured CO fluxes and the MW conversion fac- yet inside the old stellar disc could be good examples of this process. tor result in M(H2)observed/M(H2)expected of 0.48, 1.25, 0.97 and 1.19 Our SMA data suggest that these galaxies are close to or ap- for NGC 4330, NGC 4402, NGC 4522 and NGC 4569, respectively. proaching peak pressure (Kenney et al. 2004; Crowl et al. 2005; Considering that the sample of Fumagalli et al. (2009) does not in- Chung et al. 2007; Vollmer 2009;Abramsonetal.2011; Vollmer clude galaxies as faint as NGC 4330 or NGC 4522, the expected H2 et al. 2012a; Abramson & Kenney 2014), which is also supported mass could have been overestimated for these two galaxies. This by the results of simulations in the cases of NGC 4330 and NGC supports that the molecular gas mass does not necessarily change 4522 (Vollmer et al. 2006, 2012b; Vollmer 2009). Therefore, they
MNRAS 466, 1382–1398 (2017) 1394 B. Lee et al. will experience strong(er) ram pressure for a while, losing more H I. galaxies that we observed using the SMA are expected to go through As long as H I is supplied into the disc, the galaxy should be capable a similar evolutionary sequence and will be found with a truncated of forming molecular gas, keeping up with star formation. Hence, FUV disc of a comparable size to the Hα and CO disc within their within one galaxy where star formation can be regulated, the FUV old stellar disc. Judging from the star formation quenching time- morphology should be more or less in good agreement with those scales by Crowl & Kenney (2008) and simulations by Vollmer et al. of Hα, which traces more recent star formation on a time-scale of (2008, 2012a), the time it will take for our SMA sample to become <20 Myr (Kennicutt 1998). This is indeed true for isolated galaxies a galaxy like NGC 4569 is expected to be only a few hundreds Myr. such as the sample of Leroy et al. (2008). In addition, the FUV extent in isolated galaxies is usually smaller than the H I extent as 5.3.2 Extraplanar space shown in the comparison of radial profiles (Leroy et al. 2008). In our galaxies, however, the Hα looks much smaller than the FUV disc, While a galaxy is undergoing very strong ram pressure, some molec- and quite distinct in morphology. This implies that star-forming disc ular gas might be pushed away from the stellar disc as clearly seen in these galaxies has been shrunken at least in the last 100 Myr. in NGC 4522. In order to test whether dense molecular gas can get While the outer disc undergoes star formation truncation, the stripped from a galaxy or not, we compare the ram pressure acting Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021 inner disc is also less likely to form stars in the same manner as on our sample due to the ICM wind with the restoring force (per field galaxies, despite enough molecular gas present. In the upstream unit area) by the galactic potential as a function of galactic radii ρ V 2 V 2 R−1 side, molecular gas is found to be compressed as if it is pushed by (i.e. ICM gal versus rot gas) in Fig. 8. The ICM density and the ICM pressure. This can be the result of direct molecular gas the orbital velocity of each galaxy are adopted from the model of compression but also molecular gas excess due to formation from the Virgo cluster (Vollmer et al. 2001; Yoon 2013). The shaded compressed diffuse gas (Henderson & Bekki 2016). Consequently, region indicates the range of possible orbital velocities at a given the star formation rate is likely to be locally increased as Fujita projected clustercentric distance. To estimate the restoring force at & Nagashima (1999) suggested (see also Kronberger et al. 2008; given galactic radii, we take the mean molecular gas and the H I gas Bekki 2014). Indeed, we find strong FUV enhancement on the same surface densities along the major axis from the radial profiles (i.e. side where the outer CO disc is found to be compressed. Figs 3c–5c) using our SMA CO data and the H I data (VIVA; Chung Alternatively, this might be the result of the removal of dust due et al. 2009). Hence, in this calculation, we do not include the IRAM to the ICM pressure and hence the upstream side being unveiled. data of NGC 4522, and therefore the restoring force of this case However, the FUV disc also shows morphological peculiarities and in Fig. 8 is only the lower limit of the lowest possible value. The it is quite distinct from the old stellar disc (e.g. a tail in NGC 4330, parameters used to estimate the restoring force and ram pressure different extents and scale heights of NGC 4402, extraplanar FUV are summarized in Table 4. emission in NGC 4522), which cannot be explained by dust strip- Our rough estimate suggests that it is not impossible to remove ping alone. In addition, more evidence for increased star formation the molecular gas in the very outer region in some circumstances. due to gas compression by the ICM pressure has been found in other However, by adopting the rotational velocity of the CO disc, which galaxies (e.g. Merluzzi et al. 2013; Chung & Kim 2014), which also in all cases has a radius less than 5 kpc, we are focused on a re- seems to be the case in our sample. Local FUV enhancements found gion whose potential is dominated by the stellar component, not near CO compression within the old stellar disc (such as the north- the dark matter halo. Therefore, in practice, the actual stripping east of NGC 4330, the east of NGC 4402 and the south-east of of the molecular gas must be more difficult than our estimation. NGC 4522) can be good examples of induced star formation by Also, depending on the encounter angle between the ICM wind and ram pressure (Figs 3e–5e and Figs 6a–c). the galactic disc, the inner molecular gas can be better shielded, Some interstellar gas removed from the main stellar disc yet which will make the stripping even harder. Lastly, considering that lingering around in a galactic halo may be re-accreted on to the disc the ISM is one entity after all and the total gas surface density after core crossing. NGC 4569 shown in Fig. 6(d), is known to be of the two phases is more appropriate for more realistic estimation, such case. The single-arm structure in H I agrees well in morphology the chance for the molecular gas to get stripped can be even less. In and kinematics with a model of stripped gas falling back to the addition, even if the molecular gas is pushed off from the galactic galaxy, which assumes that the galaxy is currently moving away disc, whether the extraplanar molecular gas leaves the galactic halo from the cluster centre after crossing the core ∼a few hundreds for good or partially comes back is still debatable. Considering that Myr ago (Vollmer et al. 2004). Its star formation quenching time- cluster galaxies are not significantly different from their field coun- scale estimated by Crowl & Kenney (2008) is also somewhat longer terpart in CO luminosity (Stark et al. 1986; Kenney & Young 1989), than our SMA sample by ∼100 Myr or so, indicating that this galaxy the molecular gas completely stripped out of a galaxy is inferred to has experienced strong ICM pressure relatively longer ago than the be a very small fraction. other three galaxies. Our estimate is, however, only the prediction for how easy or What is intriguing in the case of NGC 4569 is that the extent of hard it is to remove the molecular gas from a galaxy completely, the FUV disc is comparable to that of the Hα and CO (Fig. 6d). and the molecular gas can be still disturbed without stripping, In this case, the star formation is limited in the central region of as our data clearly show. In fact, the recent result by Fumagalli the galaxy as most of the H I gas at larger galactic radii is stripped. et al. (2009) that severely H I-deficient galaxies tend to be found Some gas can be re-obtained but not enough gas is left to replenish with lower molecular gas surface density may simply reflect the the entire disc. Therefore, the FUV extent can decrease to a similar fact that the molecular gas of galaxies under strong ICM pressure size as the Hα extent because FUV emission may fade as time goes is perturbed, not necessarily implying that molecular gas being by, if molecular gas cannot be replenished because the atomic gas deficient. has been stripped. At this stage, it is similar to isolated galaxies Extraplanar molecular gas may also form new stars (e.g. ESO in a sense that FUV is overall well correlated with Hα/CO, but 137-001 with a prominent RPS tail found in a massive cluster, now the star-forming disc is somewhat shrunken in comparison to Abell 3627; Sun, Donahue & Voit 2007; Sun et al. 2010;Jachym´ what it used to be during active H I stripping or before. The three et al. 2014). The FUV emission with several Hα knots along the H I
MNRAS 466, 1382–1398 (2017) Molecular gas properties of Virgo spirals 1395 Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021
Figure 8. The gravitational restoring force acting on the molecular gas of our sample is probed as a function of galactic radius and compared with the current ram pressure estimate. The dashed and the solid lines represent the molecular and the atomic gas, respectively, with the receding side in red and the approaching side in blue. The dotted green line represents the maximum total gas surface density, i.e. (H Imax + H2max) measured at given radii. The upper arrow at the bottom of each panel corresponds to the galaxy radius measured in the R band. The vertical range in the ram pressure value (shaded area) corresponds to a range in orbital velocities of each galaxy at a given projected cluster centric distance from the model of Vollmer et al. (2001). Detailed parameters adopted in this plot can be found in Table 4.
Table 4. Parameters of the restoring force and ram pressure. in NGC 4330 and NGC 4402, but with mass smaller than our de- tectability or that of the IRAM (Vollmer et al. 2008). NGC 4330 NGC 4402 NGC 4522 Together with this, there is also an important difference be- Ram pressure parameters tween NGC 4330 and NGC 4522, which is the presence of CO ◦ a in the extraplanar star formation region. In the case of NGC 4330, dM87( ) 2.1 1.4 3.3 −5 −3 b more rapidly increasing ram pressure compared to NGC 4522 might ρICM(10 cm ) 8.2 63.8 20.4 −1 c be responsible for faster dissociation of molecular cloud, and hence Vgal(km s ) 1255–1775 1338–1892 1138–1609 absence of CO along the FUV tail. Alternatively, some fraction of Restoring force parameters stripped molecular gas might be destroyed by high-energy photon R (kpc)d 0.2–4.2 0.2–4.9 0.2–2.1 emitted from massive young stars (Crowl et al. 2005). Therefore, 19 −2 e H2 (10 cm ) 1.4–109 6.6–568 1.5–187 extraplanar or stripped CO may not be always observable even if it −1 f Vrot(km s ) 140 145 122 is pushed off from the stellar disc. aProjected distance from M87. bICM density. cRange of possible orbital velocities. 6 CONCLUSIONS dRadius from the galactic centre (minimum and maximum values from either In this work, we have presented high-resolution 12CO (2–1) data side). eMolecular gas surface density (minimum and maximum values from either obtained using the SMA of three Virgo spirals, NGC 4330, NGC side). 4402 and NGC 4522, which are undergoing active H I stripping W f V = H I,20 W due to ram pressure. By comparing with single-dish 12CO (2–1), Rotational velocity, rot 2sini , i: the inclination angle, H I,20:theHI linewidth (Chung et al. 2009). H I,FUV,Hα and DSS2 red data, we have discussed how ICM pressure changes molecular gas and star formation properties. Most important results from our study can be summarized as follows: tail of NGC 4330 and the extraplanar FUV/Hα emission in NGC 4522 might be such a case. Outside the old stellar disc, stars may (i) The overall CO morphology and kinematics are quite asym- form in two ways, from stripped molecular gas, or alternatively from metric and disturbed. Morphological peculiarities found in CO are newly formed molecular gas. Indeed, Tonnesen & Bryan (2012)in closely related to H I morphology in these galaxies, indicating that their simulations show that stripped diffuse gas can form dense the molecular gas is also affected by the ram pressure. molecular gas by cooling process, which can possibly form new (ii) We find the CO clump in the south-west of NGC 4330 stars. However, in extraplanar space, it is questionable how effi- (Figs 3a–c) and the modest CO bump in the north-west of NGC ciently stripped atomic hydrogen can be pressurized, turning into 4402 (Figs 4a–c), with distinct velocity structure from the overall molecular form (Blitz & Rosolowsky 2006). Therefore, star forma- velocity gradient of the molecular disc. This also supports the idea tion taking place outside the main disc is more likely to be utilizing that molecular gas can be displaced by the external pressure inside the molecular gas that is pushed from the main disc but has not left a stellar disc. the galactic halo. (iii) Both FUV and Hα are enhanced where H I and CO are com- As in NGC 4522, however, there are several extraplanar Hα re- pressed, supporting that ram pressure can also trigger star formation gions with no detectable CO. This suggests that the H2 gas depletion temporarily (e.g. Figs 4d–f of NGC 4402). In these galaxies, how- τ time-scale ( H2 ) in extraplanar space can be much shorter than what ever, FUV shows distinct morphology and extent from those of CO, it is on the disc (of the order of a few Gyr; Bigiel et al. 2008, 2011). while Hα and CO are overall in good agreement (Fig. 6). The large τ α α Meanwhile, H2 at the location of H /CO blob is measured to be offset between FUV and H must indicate that the star formation comparable to what is normally found in the galactic disc (see has been recently quenched, likely over the last 100 Myr as the Appendix B). Therefore, there might be some extraplanar CO even molecular gas properties have changed.
MNRAS 466, 1382–1398 (2017) 1396 B. Lee et al.
(iv) To compare with NGC 4569 (Fig. 6d), which is thought to Bohringer¨ H., Briel U. G., Schwarz R. A., Voges W., Hartner G., Trumper¨ have already crossed the cluster core, the offset between FUV and J., 1994, Nature, 368, 828 Hα/CO is expected to smooth out over time (∼a few hundreds Myr) Boselli A., Cortese L., Boquien M., Boissier S., Catinella B., Gavazzi G., after the core crossing as the galaxy further loses H I, i.e. the capa- Lagos C., Saintonge A., 2014, A&A, 564, A67 bility of forming molecular gas, and star formation is completely Cayatte V., van Gorkom J. H., Balkowski C., Kotanyi C., 1990, AJ, 100, 604 Chung E. J., 2012, PhD thesis, Yonsei University shut down at large galactic radii. Chung E. J., Kim S., 2014, PASJ, 66, 11 In summary, our results suggest that a strong ICM pressure not Chung A., van Gorkom J. H., Kenney J. D. P., Vollmer B., 2007, ApJ, 659, only strips diffuse atomic gas but also changes the properties of L115 dense molecular gas in the inner few kpc of a galaxy. Molecular Chung A., van Gorkom J. H., Kenney J. D. P., Crowl H., Vollmer B., 2009, AJ, 138, 1741 gas can be pushed from the stellar disc in the outer part as reported Crowl H. H., Kenney J. D. P., 2008, AJ, 136, 1623 in previous studies, yet we find that the molecular gas is less likely Crowl H. H., Kenney J. D. P., van Gorkom J. H., Vollmer B., 2005, AJ, 130, to be completely stripped from a galaxy. On the side where the 65
ICM pressure acts, molecular gas can be enhanced by compression Davies R. D., Lewis B. M., 1973, MNRAS, 165, 231 Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021 and molecular formation from atomic gas, locally increasing star Drimmel R., Spergel D. N., 2001, ApJ, 556, 181 formation. As the gas is rapidly used up, the star-forming disc Fujita Y., Nagashima M., 1999, ApJ, 516, 619 will decrease in size as no sufficient atomic gas is left to form Fumagalli M., Krumholz M. R., Prochaska J. X., Gavazzi G., Boselli A., more molecular gas. On the opposite side, molecular gas can be 2009, ApJ, 697, 1811 perturbed in a way that reflects the atomic gas, which may prevent Gavazzi G., Boselli A., Donati A., Franzetti P., Scodeggio M., 2003, A&A, a galaxy from forming stars as before, i.e. injecting turbulence into 400, 451 Gil de Paz A. et al., 2007, ApJS, 173, 185 molecular gas is likely to cause star formation rates to be inefficient Giovanelli R., Haynes M. P., 1985, ApJ, 292, 404 (Alatalo et al. 2015). The combination of these two effects must be Gunn J. E., Gott J. R., III1972, ApJ, 176, 1 an important driver that makes galaxies evolve to become passive Haynes M. P., Giovanelli R., 1984, AJ, 89, 758 in the cluster environment. Helfer T. T., Thornley M. D., Regan M. W., Wong T., Sheth K., Vogel S. N., In order to investigate the molecular gas properties and its con- Blitz L., Bock D. C.-J., 2003, ApJS, 145, 259 ditions for star formation under strong ICM pressure more quan- Henderson B., Bekki K., 2016, ApJ, 822, L33 titatively, we are currently probing the density and temperature of Hidaka M., Sofue Y., 2002, PASJ, 54, 33 molecular gas using multitransitions and isotopes of CO (Lee et al. Jachym´ P., Combes F., Cortese L., Sun M., Kenney J. D. P., 2014, ApJ, 792, in preparation). Our future work is expected to provide further in- 11 sights on the evolution of molecular gas and thus star formation Kenney J. D. P., Young J. S., 1989, ApJ, 344, 171 Kenney J. D. P., Young J. S., Hasegawa T., Nakai N., 1990, ApJ, 353, 460 efficiency in galaxies undergoing ram pressure. Kenney J. D. P., van Gorkom J. H., Vollmer B., 2004, AJ, 127, 3361 Kenney J. D. P., Abramson A., Bravo-Alfaro H., 2015, AJ, 150, 59 ACKNOWLEDGEMENTS Kennicutt R. C., Jr, 1998, ARA&A, 36, 189 Koopmann R. A., Kenney J. D. P., 2004a, ApJ, 613, 851 Support for this work was provided by the National Research Koopmann R. A., Kenney J. D. P., 2004b, ApJ, 613, 866 Foundation of Korea to the Center for Galaxy Evolution Research Koopmann R. A., Kenney J. D. P., Young J., 2001, ApJS, 135, 125 (No. 2010-0027910) and Science Fellowship of POSCO TJ Park Kregel M., van der Kruit P. C., de Blok W. J. G., 2004, MNRAS, 352, Foundation. This work has been also supported by NRF grant no. 768 2015R1D1A1A01060516. The SMA is a joint project between the Kronberger T., Kapferer W., Ferrari C., Unterguggenberger S., Schindler S., 2008, A&A, 481, 337 Smithsonian Astrophysical Observatory and the Academia Sinica Lee B., Chung A., 2015, Publ. Korean Astron. Soc., 30, 491 Institute of Astronomy and Astrophysics and is funded by the Smith- Leroy A. K., Walter F., Brinks E., Bigiel F., de Blok W. J. G., Madore B., sonian Institution and the Academia Sinica. This work made the Thornley M. D., 2008, AJ, 136, 2782 use of HERACLES, ‘The HERA CO-Line Extragalactic Survey’ Leroy A. K. et al., 2009, AJ, 137, 4670 (Leroy et al. 2009). We acknowledge the usage of the Hyper- Lucy L. B., 1974, AJ, 79, 745 Leda database (http://leda.univ-lyon1.fr). This research has made Mei S. et al., 2007, ApJ, 655, 144 use of the GOLDMine Database (Gavazzi et al. 2003), and it Merluzzi P. et al., 2013, MNRAS, 429, 1747 made use of APLPY, an open-source plotting package for PYTHON Mok A. et al., 2016, MNRAS, 456, 4384 hosted at http://aplpy.github.com. ST was supported by the Alvin E. Nehlig F., Vollmer B., Braine J., 2016, A&A, 587, A108 Nashman Fellowship in Theoretical Astrophysics. BL is grateful to Paturel G., Petit C., Prugniel P., Theureau G., Rousseau J., Brouty M., Dubois P., Cambresy´ L., 2003, A&A, 412, 45 E.-J. Chung for the technical help with GIPSY. Qi C., 2012, Available at: http://www.cfa.harvard.edu/∼cqi/mircook.html Rengarajan T. N., Iyengar K. V. K., 1992, MNRAS, 259, 559 REFERENCES Rhee M.-H., van Albada T. S., 1996, A&AS, 115, 407 Salo H. et al., 2015, ApJS, 219, 4 Abramson A., Kenney J. D. P., 2014, AJ, 147, 63 Skrutskie M. F. et al., 2006, AJ, 131, 1163 Abramson A., Kenney J. D. P., Crowl H. H., Chung A., van Gorkom J. H., Sofue Y.,Koda J., Nakanishi H., Onodera S., Kohno K., Tomita A., Okumura Vollmer B., Schiminovich D., 2011, AJ, 141, 164 S. K., 2003a, PASJ, 55, 17 Alatalo K. et al., 2015, ApJ, 812, 117 Sofue Y., Koda J., Nakanishi H., Onodera S., 2003b, PASJ, 55, 59 Barmby P. et al., 2006, ApJ, 650, L45 Sofue Y., Koda J., Nakanishi H., Hidaka M., 2003c, PASJ, 55, 75 Bekki K., 2014, MNRAS, 438, 444 Solomon P. M., Downes D., Radford S. J. E., Barrett J. W., 1997, ApJ, 478, Bigiel F., Leroy A., Walter F., Brinks E., de Blok W. J. G., Madore B., 144 Thornley M. D., 2008, AJ, 136, 2846 Stark A. A., Knapp G. R., Bally J., Wilson R. W., Penzias A. A., Rowe H. Bigiel F. et al., 2011, ApJ, 730, L13 E., 1986, ApJ, 310, 660 Blitz L., Rosolowsky E., 2006, ApJ, 650, 933 Strong A. W., Mattox J. R., 1996, A&A, 308, L21
MNRAS 466, 1382–1398 (2017) Molecular gas properties of Virgo spirals 1397
Sun M., Donahue M., Voit G. M., 2007, ApJ, 671, 190 Vollmer B., Wong O. I., Braine J., Chung A., Kenney J. D. P., 2012b, A&A, Sun M., Donahue M., Roediger E., Nulsen P. E. J., Voit G. M., Sarazin C., 543, A33 Forman W., Jones C., 2010, ApJ, 708, 946 Walter F., Brinks E., de Blok W. J. G., Bigiel F., Kennicutt R. C., Jr, Thornley Swaters R. A., 1999, PhD thesis, Univ. Groningen M. D., Leroy A., 2008, AJ, 136, 2563 Tonnesen S., Bryan G. L., 2012, MNRAS, 422, 1609 Warmels R. H., 1988a, A&AS, 72, 19 Verheijen M. A. W., 1997, PhD thesis, Univ. Groningen Warmels R. H., 1988b, A&AS, 72, 427 Vollmer B., 2009, A&A, 502, 427 Wong O. I., Kenney J. D. P., Murphy E. J., Helou G., 2014, ApJ, 783, 109 Vollmer B., Cayatte V., Balkowski C., Duschl W. J., 2001, ApJ, 561, 708 Yasuda N., Fukugita M., Okamura S., 1997, ApJS, 108, 417 Vollmer B., Balkowski C., Cayatte V., van Driel W., Huchtmeier W., 2004, Yoon H., 2013, MS thesis, Yonsei University A&A, 419, 35 Vollmer B., Soida M., Otmianowska-Mazur K., Kenney J. D. P.,van Gorkom J. H., Beck R., 2006, A&A, 453, 883 APPENDIX A: 12CO GAS DISTRIBUTION ON Vollmer B., Braine J., Pappalardo C., Hily-Blant P., 2008, A&A, 491, 455 Vollmer B. et al., 2012a, A&A, 537, A143 THE OPTICAL COLOUR IMAGE Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021
Figure A1. 12CO (2–1) contours (white contours: SMA, yellow contours: IRAM) are overlaid on optical colour images (NGC 4330: WIYN 3.5 m telescope BVR colour image; Abramson et al. 2011, NGC 4402 and NGC 4522: HST3 BVI colour images). Top: NGC 4330. Middle: NGC 4402. Bottom: NGC 4522. The physical scale bar (20 arcsec) of each galaxy is shown at the bottom left. The blue cross indicates the stellar disc centre of each galaxy.
3Based on observations made with the NASA/ESA Hubble Space Telescope, obtained from the data archive at the Space Telescope Science Insti- tute. STScI is operated by the Association of Universities for Research in Astronomy, Inc. under NASA contract NAS 5-26555.
MNRAS 466, 1382–1398 (2017) 1398 B. Lee et al.
IRAM 12CO (2–1) data of Vollmer et al. (2008). Using Hα image APPENDIX B: THE H GAS DEPLETION 2 (Koopmann et al. 2001), the star formation rate in each region TIME-SCALE (τ ) IN THE EXTRAPLANAR H2 (Fig. B1, red squares) is estimated as follows: REGION OF NGC 4522 SFR = 7.9 × 10−42L(H α)(B1) We estimate the molecular gas depletion time-scale (the ratio of molecular gas mass and star formation rate) in extraplanar region in M yr−1,whereL(Hα)istheHα luminosity in erg s−1 (Kenni- of NGC 4522 following the prescription below. We first calculate cutt 1998). Measured depletion time-scales are 1.1 Gyr (region 1) the molecular gas mass in two regions (Fig. B1, red squares) using and 0.8 Gyr (region 2), respectively. Downloaded from https://academic.oup.com/mnras/article/466/2/1382/2646788 by guest on 17 March 2021
Figure B1. The Hα image is shown in black contours overlaid with IRAM 12CO (2–1) image in blue contours. The beam size of the IRAM is shown in blue circle at the bottom left. Two regions with extraplanar Hα and CO emissions where the molecular gas depletion time has been estimated are indicated in red boxes.
This paper has been typeset from a TEX/LATEX file prepared by the author.
MNRAS 466, 1382–1398 (2017)